Bruxism, a non-functional gnashing or grinding of teeth, may be viewed as a learned response to arousal. The psychological treatments derived from this learning-theory viewpoint have shown promise in controlling this behavior, but their effectiveness has not been evaluated in controlled, clinical outcome trials. This project proposes to develop and evaluate a number of behaviorally-based psychological treatments for bruxism. Bruxing patients will be given either a classical conditioning treatment for nocturnal bruxism or a biofeedback treatment for diurnal clenching. The classical conditioning treatment involves all-night studies in which physiological responses of the subject are programmed so that a tone sounds when an accident of bruxism occurs. Thus, bruxism becmes a conditioned stimulus which is followed by a tone (UCS), awakening (UCR), and inhibition of grinding (CR). It is hypothesized that subjects will learn to sleep through the night without grinding. Subjects will receive contingent and non-contingent treatment in a counter-balanced design. In the biofeedback study, subjects will be given either masseter EMG biofeedback, false feedback or be asked to self-monitor clenching and grinding. A study designed to investigate ways of increasing generalization effects from biofeedback treatments is also planned.